Ally Financial empowers customer service associates to focus on human engagement by using Azure OpenAI Service

Improving AI with Customer Feedback

After evaluating several options, Ally launched a collaborative project to develop a pilot built on Microsoft Azure using Azure OpenAI Service. It sought to complete the pilot within six weeks and have the first use case in eight, an exceptionally short timeline for the banking sector. By incorporating feedback from customer service associates, Ally addressed employees’ questions about AI and improved the new solution.  Now more than 700 associates have real-time access to summaries of tens of thousands of customer service calls per week and can edit them as needed. The integration with AI empowers associates to focus on answering customer questions and saves them valuable minutes in servicing customer inquiries. 

“It’s been very exciting to watch this technology empower our associates and give them a better tool to serve customers,” says Diane Morais, President of Consumer and Commercial Banking at Ally. 

Delivering a Generative AI Solution in Eight Weeks with Azure OpenAI Service

Ally’s customer experience associates provide the human touch that customers expect. When the firm saw an opportunity to use generative AI to help it deliver enhanced value and personal service, it evaluated many vendor offerings. Ally found that Azure OpenAI Service was the best solution for its use case. “The massive investment Microsoft has made in OpenAI helped move things forward, but the key thing is that Microsoft really understands our need to implement robust, reliable, and secure capabilities,” says Scott Richardson, Chief Information Officer of Enterprise Data and Analytics at Ally.

To deliver value quickly, Ally set up a collaborative development effort. Business, technical, and customer service teams worked together to define needs, provide feedback, and fine-tune the AI. Ally identified three keys to successful development. First, it would learn from employees before implementing new technology for customers. Second, Ally would keep a human in the middle of all AI interactions and have oversight of all outputs. Finally, the company’s data would stay within its own firewalls, so the foundational model would not learn from Ally’s data. Following these guidelines, and with support from Microsoft, Ally’s technology group delivered a production-ready solution in just eight weeks. 

Targeting a 50 % Reduction in Post-call Effort with Conversational AI

Ally customer service associates no longer take extensive notes while interacting with a customer or write detailed call summaries afterward. With the conversational AI capabilities of Azure OpenAI Service, they simply review an automatically generated summary and confirm that it contains all the necessary information. The new solution has cut associates’ post-call effort by 30 percent—with the expectation of a 50 percent reduction—while capturing key details automatically. “Our associates say that as they read through the summaries, they often find information from the call that they had forgotten,” says Wendy Dempsey, Executive Director of Customer Experience at Ally.

Ally leaders attribute the accuracy of its solution to collaboration between its technology and customer service teams. The company kept things simple: after viewing the AI-generated call summary, associates gave either a thumbs-up or a thumbs-down rating. Based on their feedback, the technology team fine-tuned its models and prompts to deliver better results. Today, accuracy is over 85 percent, and Ally is still improving the solution. 

AI in the Call Center

Ally also took a proactive approach to addressing worries about AI in the call center. When employees voiced concerns about losing work, Ally assured them that the technology would be empowering. “The goal was to free up associates to truly focus on the customer and support that customer-obsessed mentality they have,” says Dempsey. “When they saw how beneficial the call summaries were, it alleviated their concerns.” Call-center associates are now enjoying a more meaningful work experience with less paperwork and more human interaction.

Ally teams are proud of the work they have done to tackle an industry-wide challenge. Banking and financial services are highly regulated, but Ally delivered a secure, responsible, and effective AI solution in an exceptionally short time frame. The firm emphasizes the importance of behind-the-scenes work in accelerating delivery. “If we didn’t have our data in a centralized place, we would not be able to take advantage of generative AI,” says Muthukrishnan. “It’s the work that goes on behind the scenes, where you invest patience, money, technology, and strategic thinking, that pays off when you launch a solution like our Ally.ai platform alongside a partner like Microsoft.”

Advancing the future of secure AI with Microsoft

Ally remains focused on its customers and creating and capturing value through innovation to better serve them. Its technology organization has the ambitious goal of launching a new project each month as it looks to harness more new uses for AI. Ally is also committed to protecting data privacy and security throughout its journey with AI, and it collaborates with Microsoft to advance that goal. “Microsoft understands our desire for responsible, secure AI,” says Richardson. “They addressed our major concerns in advance, which made it easy to collaborate with them.”

Ultimately, Ally’s embrace of AI is about empowering employees and serving its customers. “For us, it’s about a relentless focus on our customers’ financial well-being and supporting our employees who are interacting with them every day,” says Muthukrishnan. “Microsoft understands these core principles and has helped us make our associates’ work experience more efficient and productive, so they can continue to serve customers at the highest level.”